Food Insecurity, Dietary Diversity, and Coping Strategies in Jordan during the COVID-19 Pandemic: A Cross-Sectional Study

The 2019 coronavirus disease (COVID-19) is an emerging respiratory infection with severe impacts on the nutritional status of the worldwide population. This cross-sectional study was conducted to assess the food insecurity, dietary diversity, and food-related coping strategies in Jordan during the pandemic using an online, self-administered questionnaire. Among the 740 adults who completed the survey, the prevalence of food security was 84.1%, whereas 2% and 13.9% were moderately and severely food-insecure, respectively. The determinants of food insecurity were educational level, monthly income, marital status, availability of health insurance, and type of residence. In addition, food insecurity was significantly higher among the participants who consumed two or fewer meals per day (p = 0.015). Moreover, an acceptable food consumption score was shown among 76.2% of the participants, and the remaining participants were at borderline (14.1%) or had poor scores (9.7%), with a significant association between food insecurity and food consumption scores (p < 0.001). The food-related coping strategies studied were significantly associated with food insecurity at both levels (p < 0.001) and were more evident in the severely food-insecure group. These findings highlight the adverse effects of COVID-19 restrictions on nutritional status, especially among food-insecure households, which might reduce food accessibility due to economic difficulties.


Introduction
In December 2019, COVID-19, a disease caused by the SARS-CoV-2 virus, was discovered, which can lead to mild-moderate respiratory symptoms [1,2]. In most symptomatic infected people, the symptoms start within less than a week and include fever, cough, fatigue, loss of taste or smell, headache, dizziness, and generalized weakness [3,4]. Generally, COVID-19 patients can recover without special medical treatment; however, older adults and groups with certain medical conditions such as respiratory diseases, cardiovascular diseases, cancer, and diabetes may develop severe symptoms and require hospitalization [5]. According to WHO statistics as of 22 March 2022, the worldwide confirmed cases

Study Design
This was a cross-sectional study conducted in Jordan. The validated Arab Food Security Scale survey questionnaire [21] was used to collect the data (Supplementary). The data were collected using a self-administered online survey. A non-probability sample of social media users on Facebook, Messenger, and WhatsApp was recruited by conducting an advertisement linked to Facebook and WhatsApp. Recruitment occurred from 28 September to 8 December 2021, and the advertisement targeted adults aged ≥ 18 years residing in Jordan.

Ethical Approval
The study was proved by the Institutional Review Board at the Hashemite University (IRB No. 8/13/2020/2021).

Survey Tool
The questionnaire involved an introduction showing the study's objectives, which highlighted that the answers of participants would be treated confidentially. The participants were not paid for their involvement in this study and it was entirely voluntary. The completion of the online survey took approximately 8-10 min. The data were presented as multiple-choice questions in a structured survey questionnaire. The first section determined participants' socio-demographic variables, including gender, marital status, age, education level, employment status, number of children, urban/rural residence, number of household members, and monthly household income. The next sections assessed Jordan's food security, food consumption, and coping strategies during the COVID-19 pandemic.

Sample Size
The minimum sample size needed to estimate the prevalence of food insecurity within a margin of error of 5% at a significance level of 5% and a power of 80% was 384. The sample size was increased to have a higher power.

Data Classification
Food insecurity measures the severity of food insecurity based on people's responses to questions about the constraints on their ability to obtain adequate food. It was assessed using the eight-item scale by adopting the Arabic version of The Food Insecurity Experience Scale (FIES) that was developed by the Food and Agriculture Organization (FAO) [22]. Participants were asked about food insecurity concerns over the past six months-the scores in this survey range from 0 to 1. Respondents answered yes/no to the eight questions, and the responses were aggregated to give raw scores ranging from 0 to 8. Food insecurity (FI) was divided into three categories, including: (1) food-secure (FS) with raw scores = 0-3; (2) moderate food insecurity (MFI), with raw scores = 4-6; and (3) severe food insecurity (SFI), with raw scores = 7-8 [23].
The Food Consumption Score (FCS) was based on a household's frequency of consumption of various food groups over the past seven days. A household food consumption score was calculated by multiplying each food group frequency by each food group weight (Table 1), and then summing these scores into one composite score [24].
A household score was compared with pre-established thresholds that indicated that household's food consumption status. The World Food Program (WFP) finds the following thresholds applicable in a wide range of situations: poor food consumption: 0 to 21; borderline food consumption: 21.5 to 35; and acceptable food consumption: > 35. A household score could have a maximum value of 112, which implied that each food group was consumed every day for the last seven days. Multinomial logistic regression was used in estimating the association between the selected household characteristics and food security status. In this case, households were classified as moderately food-insecure and severely food-insecure. A p-value of < 0.05 was considered statistically significant.

Socio-Demographic Characteristics
A total of 740 Jordanian adults participated in this study; of them, 160 (21.6%) were males and 580 (78.4%) females. Their socio-demographic characteristics classified according to food security status are listed in Table 2. The majority of the participants were younger than 50 years (92.7%), had a diploma or bachelor's at an educational level (67.8%), lived in the city (85.4%), lived in their own residence (70.7%), and had health insurance (66.4%). More than half of our study population was single (55%) and had a family with 4-7 members (52.4%). Further, about 45% of the participants had a household income of between JOD 500 and 1000 (USD 700-1400).

Household Food Security Status
The prevalence of food security among the study population was 84.1%, while 2% were classified as moderately food-insecure and about 14% as severely food-insecure, as shown in Table 2. Associations between socio-demographic characteristics and food insecurity are also presented in Table 2. A significant association was found between the presence of food insecurity and having a low educational level, being divorced or widowed, having more than three children, having a maximum of one employed family member, living in a rented residence, not having health insurance, and an overall household income of lower than JOD 500 (p < 0.05).

Determinants of Household Food Insecurity
As many socio-demographic factors are significantly associated with food security, a multinomial logistic regression analysis was performed to evaluate the strength of the relationship between these variables and the risk of being food-insecure at both levels, moderate or severe, as shown in Table 3. Educational level, marital status, and income level were all determinants of being severely food-insecure, while the unavailability of health insurance was the determinant that could lead to being moderately food-insecure. The results showed that the participants with a low educational level were more likely to be severely insecure by about seven times (95% CI = 1.104-3.978; p < 0.001) for those with secondary school or less and two times (95% CI = 3.401-15.881; p < 0.05) for those with a diploma or bachelor's degree when compared with postgraduate participants. Participants with a monthly income of less than JOD 500 were about seven times more likely to be severely food-insecure when compared with participants achieving more than JOD 1000 (95% CI =3.336-13.651; p < 0.001). Single participants less likely to be severely food-insecure, with an OR = 0.219 (95% CI = 0.077-0.622; p < 0.05). Further, the availability of health insurance was found to be a helpful factor that protected participants from being moderately food-insecure (OR (95% CI) = 0.167 (0.053-0.532); p < 0.05). Moreover, participants who were living in their own residences tended to be at a lower risk of being food-insecure at both levels (moderately food-insecure, OR = 0.308 (95% CI = 0.110-0.863; p < 0.05), and severely food-insecure, OR = 0.420 (95% CI = 0.274-0.643; p < 0.05)) ( Table 3).

Weekly Food Group Consumption
Approximately 63% of the participants consumed two or fewer meals daily, and a higher proportion of them consumed almost all of the food groups in a frequency of 3 days or fewer per week, as shown in Table 4. Food insecurity was significantly higher among the participants who consumed two or fewer meals (3% and 15.3% vs. 0.4% and 11.6% (p = 0.015) for moderately and severely food-insecure; respectively) compared with those who consumed more meals. Furthermore, a higher prevalence of food insecurity was significantly shown for participants who reported that they consumed meals less than usual (about 5% moderate food insecurity and 29% severe food insecurity) (p < 0.001). With respect to food group weekly consumption, all the main food groups were significantly associated with food security status. Participants who consumed cereals, vegetables, fruits, eggs, legumes and nuts, milk and dairy products, oil and fat, sugar, spices and condiments, and meat and poultry at a frequency of 3 days or fewer per week tended to be more food-insecure than participants who consumed these items more during the week (Table 4).

Food Consumption Scores
The weekly food group consumption was translated into food consumption scores and classified into three groups: poor (9.7%), borderline (14.1), and the remaining (76.2%) were classified as consuming food at an acceptable level. There was a strong significant association between food insecurity, at both levels of severity, and the food consumption scores, as shown in Figure 1 (p < 0.001). The associations were stronger with the more severe range of food insecurity. These findings were also reflected in poor food consumption scores, which were significantly lower among food-secure households (Figure 1). and classified into three groups: poor (9.7%), borderline (14.1), and the remaining (76.2%) were classified as consuming food at an acceptable level. There was a strong significant association between food insecurity, at both levels of severity, and the food consumption scores, as shown in Figure 1 (p < 0.001). The associations were stronger with the more severe range of food insecurity. These findings were also reflected in poor food consumption scores, which were significantly lower among food-secure households ( Figure 1).

Food-Based Coping Strategies during the Last Week
During the last week of response, more than 96% of the participants were following food-based coping strategies for three days per week or less. Participants reported that they were eating cheaper food, borrowed food, and fewer meals or small amounts of food, and there were even adults eating less food to spare it for their children ( Table 5). The associations between food-based coping strategies and food security status are presented in Table 5. Almost all the coping strategies were significantly associated with food insecurity at both levels (p < 0.001), but this association was evident in the severely insecure group. Participants with severe food insecurity tended to practice these coping

Food-Based Coping Strategies during the Last Week
During the last week of response, more than 96% of the participants were following food-based coping strategies for three days per week or less. Participants reported that they were eating cheaper food, borrowed food, and fewer meals or small amounts of food, and there were even adults eating less food to spare it for their children ( Table 5). The associations between food-based coping strategies and food security status are presented in Table 5. Almost all the coping strategies were significantly associated with food insecurity at both levels (p < 0.001), but this association was evident in the severely insecure group. Participants with severe food insecurity tended to practice these coping strategies at least four days during the week, and the results revealed that 45.5% of those participants ate cheaper food, 25% borrowed food, 36.4% ate fewer meals, 39.1% ate a small amount of food, and 13.9 % were instances where only adults ate less, compared to food-secure participants. In a separate analysis for non-food-based coping strategies during the last 30 days, approximately 34% of the participants did not use any strategy to meet basic food needs while 28% spent money savings to meet basic food needs, and 16% spent less money on other needs such as education and health. The rest of the participants used different strategies, such as buying food on credit, borrowing money, or doing any labor for food (results not shown).

Food Quality and Quantity
The food quality and quantity were affected by the level of food insecurity. Almost 95% of the food-secure respondents had good qualities and quantities of food, while 68% of the severe food-insecure participants had smaller quantities of food and 21% had a good quality of food ( Figure 2).

Discussion
This cross-sectional study found that most of the studied population was food-secure (84.1%), whereas 2% were moderately food-insecure and 13.9% were severely foodinsecure. These results differ from a previous study conducted in Jordan during the COVID-19 quarantine (n = 3129), which found that the prevalence of food security was 40.7%, moderate food insecurity was 36.1%, and severe food insecurity was 23.1% [25]. Compared with Elsahoryi et al.'s results [25], the improvement in food security status is expected due to the demise of some negative factors associated with the quarantine, such as the lockdowns and movement difficulty; thus, the limitations in food accessibility were reduced. In the present study, there is a significant impact of educational level on food security wherein participants with low educational levels were more likely to be severely food-insecure. This association is consistent with several past studies conducted in both rural and urban areas [26][27][28].
Education plays an important role in employment opportunities, working efficiency, accessing information about health and nutrition, increasing income, and diversity, all of which improve household food supply [29,30]. Our results showed that being divorced or widowed is another significant factor associated with food insecurity, and that being single decreases the risk of severe food insecurity. The association between marital status and household food security varies in the literature. For example, studies of households in Kenya and Nigeria (n= 209 and 180, respectively) reported that married respondents

Discussion
This cross-sectional study found that most of the studied population was food-secure (84.1%), whereas 2% were moderately food-insecure and 13.9% were severely food-insecure. These results differ from a previous study conducted in Jordan during the COVID-19 quarantine (n = 3129), which found that the prevalence of food security was 40.7%, moderate food insecurity was 36.1%, and severe food insecurity was 23.1% [25]. Compared with Elsahoryi et al.'s results [25], the improvement in food security status is expected due to the demise of some negative factors associated with the quarantine, such as the lockdowns and movement difficulty; thus, the limitations in food accessibility were reduced. In the present study, there is a significant impact of educational level on food security wherein participants with low educational levels were more likely to be severely food-insecure. This association is consistent with several past studies conducted in both rural and urban areas [26][27][28].
Education plays an important role in employment opportunities, working efficiency, accessing information about health and nutrition, increasing income, and diversity, all of which improve household food supply [29,30]. Our results showed that being divorced or widowed is another significant factor associated with food insecurity, and that being single decreases the risk of severe food insecurity. The association between marital status and household food security varies in the literature. For example, studies of households in Kenya and Nigeria (n = 209 and 180, respectively) reported that married respondents were more likely to be food-secure due to the contribution of both the husband and wife to family income, which in turn improved food security status [31,32]. The average monthly wage income of professional employees in Jordan in 2019 was JOD 692 (USD 970) [33]. On the contrary, a cross-sectional study in Iran (n = 1385) documented that married respondents were less likely to be food-secure [34]. A similar finding was reported in a study of 100 households in Ghana, and this could be attributed to the smaller number of family members, which may reduce a family's economic burden [35]. The present study found that having more than three children negatively impacted food security. The larger family size increased the competition for the available resources. A higher number of children (an inactive age group) was associated with less production and more consumption [36].
Furthermore, our results found a significant association between food insecurity and a maximum of one working member in the household. During the COVID-19 pandemic, 7% of Jordanians lost their jobs and the unemployment rate reached more than 23% in 2021 compared to approximately 19% in 2019 [37,38]. Loopstra and Tarasuk [39] reported that a change in the number of full-time employed household members was associated with the severity of food insecurity independent of the change in household income, and they explained this association by improving access to dental insurance and prescription drugs, which enabled families to save money for food. This may explain our results regarding the association between having health insurance and food security, where the unavailability of health insurance was a determinant of moderate food insecurity. Further, the relationship between the number of employed household members and food security may be related to the large effect of the number of employees on family income, as employment is one of the main sources of total income [40]. Indeed, the COVID-19 pandemic had a critical influence on households' incomes in different countries. For example, a study of 1821 Nigerian households showed that 79% reported a decline in household income during the pandemic restrictions [41]. The current study found that monthly income level is one of the determinants of severe food insecurity. Participants with a monthly income of less than JOD 500 were about seven times more likely to be severely food-insecure than participants with incomes of more than JOD 1000. In line with our results, Elsahoryi et al.'s cross-sectional study in Jordan reported that participants with a per capita monthly income of below the poverty line were more susceptible to moderate and severe food insecurity during the quarantine period in 2020 [25]. Similarly, the association between low-income levels and food insecurity has been documented in several countries such as Turkey [42], South Africa [43], Sudan [44], Bangladesh [45], Nigeria [41], and Ethiopia [46].
Regarding the type of residence, this study found that living in one's own residence decreases the likelihood of moderate and severe food insecurity compared to living in a rented residence. Studies have reported that housing costs reduce food access and, in turn, disturb the food security status of low-income families [47]. Further, and consistent with our results, people who lived in rented houses were at a higher risk of severe food insecurity than people who lived in owned houses during the COVID-19 quarantine in Jordan [25].
The current study indicates that around 63% of the participants consumed two or fewer meals per day, and this frequency was reported "as usual" by 87% of the participants. The consumption of 1-2 meals per day was higher before the COVID-19 pandemic and found to be 85% [48]. Only 2.3% of the participants in the current study reported an increase in the number of meals than usual. Higher percentages were reported in an earlier study of 4473 participants in Jordan during the lockdown, where 69.4%, 89.9%, and 54% of the sample reported an increase in breakfast, lunch, and dinner numbers, respectively [8]. The effect of the COVID-19 pandemic on the number of daily meals consumed was studied in different countries; a cross-sectional study of 2970 adults residing in the Middle East and North Africa regions found that 6.2% were consuming five or more meals per day during the COVID-19 pandemic compared to 2.2% pre-COVID-19, and 45.1% were skipping meals during the pandemic compared to 64.4% before [49]. Likewise, a study of 1012 adult residents of the United Arab Emirates compared the number of daily meals consumed before lockdown and during the lockdown and reported an increase in the consumption of 3-4 meals per day from 51.5% to 56.5%, and an increase of ≥ 5 meals per day consumption from 2.1% to 7% [50]. Further, a study of 866 university students in Turkey found that 23.0% of the participants increased the number of their daily meals during the pandemic compared to before [51]. Similarly, a study of 1932 Italians showed that 46.1% of the respondents consumed more than usual during the lockdown [52]. Cross-sectional studies from other countries also reported more meals consumed during the lockdown than before, such as India [53] and Saudi Arabia [54]. These findings show how the pandemic and its associated lockdowns have affected dietary patterns, which could be due to a variety of factors, including stress, anxiety, boredom, and the longer time available for eating during quarantine compared to working and studying outside the home, and food security status [55][56][57]. Certainly, our results showed that moderate and severe food insecurity occurs at a significantly higher prevalence among participants who consumed two or fewer meals and reported a "less than usual" number of meals compared to those who consumed more than two meals and reported a "more than usual" or "as usual" number of meals. This is in agreement with a study in Saudi Arabia (n = 879), which indicated that changes in the number of meals per day during the COVID-19 curfew were frequent among participants with severe food insecurity. These changes occurred mainly due to food unavailability [58].
In addition, the present study showed that most participants consumed almost all the main food groups 3 days or less per week. A significant association was shown between food insecurity and a frequency of 3 days or fewer per week consumption of cereals, vegetables, fruits, eggs, legumes and nuts, milk and dairy products, oil and fat, sugar, spices and condiments, meat, and poultry. When translated to food consumption scores, the current results report that 76.2% of the participants were within acceptable levels, 14.1% were classified as borderline, and 9.7% were within poor consumption levels, with a significant association between food insecurity and food consumption scores. Further, consuming good quality and quantities of food was only shown in 4.3% of food-insecure respondents. The food consumption score was studied among other countries in the region and found to have worsened during the COVID-19 pandemic; for example, a cross-sectional study of 2282 Lebanese participants reported a decrease in the average food consumption score by 4.6% during the lockdown [59]. These results could be due to the decrease in households' food purchasing power during the pandemic, which was more prevalent among food-insecure households, leading to a shortage of food access and thus decreasing food quantity and quality. However, Elsahoryi et al.'s study during the quarantine in Jordan showed no associations between weekly food group intake and food security status, except for the carbohydrate (higher intake among the severe food-insecure group) and meat groups (higher intake among the food-secure group) [25]. This further highlights the influence of economic status on the variety of foods consumed, as the price of food items from the carbohydrates group is lower than the meat, poultry, and fish group.
Most of our study respondents (96%) reported using food-based coping strategies three days per week or less. Significant associations were found between food insecurity and eating cheaper foods, eating fewer meals or small amounts of food, and adults eating less food to spare it for their children. However, this association was evident in the severely insecure group as they followed these coping strategies at least four days during the week. The world food program reported an increase in the percentage of Jordanian households following consumption-based coping strategies from 23% in 2014 to 55% during the pandemic, with 25% reducing the number of meals/day and 10% relying on friends and family for help with food shortages. This occurred mainly due to a loss of employment and income [38]. Consistent with these results, the data from a multi-country survey of 1075 participants showed that 49.2% relied on less-preferred food during the pandemic, 30.3% reduced portion size, and 25.7% reduced the number of meals. In addition, they used other coping strategies, such as restricting food for adults (18.7%) and borrowing food (17%) [60]. Furthermore, the association between food insecurity and food-based coping strategies was studied in food-insecure households in Ethiopia (n = 6671), and it showed that most households responded to food insecurity by managing food left in their homes. Decreasing the amount and frequency of meals was the most prevalent coping strategy (55.96%), followed by borrowing food and money (38.11%) [61]. Similarly, a study of Vermont households (n = 3219) revealed that food-insecure participants followed food-related coping strategies significantly more than food-secure households. This result applied to families who were food-insecure before the pandemic, as well as families who became food-insecure after the pandemic [62]. It is obvious that implementing food-related coping strategies is common for dealing with food insufficiency, which highlights the seriousness of the problem, especially among food-insecure households in the context of the COVID-19 pandemic.
Compared with the results presented in the current study, the previous studies in Jordan showed a lower prevalence of food insecurity before the COVID-19 pandemic. In 2013-2014, a study of 24,740 households reported that 93.8% of the sample were food-secure and 0.5% were food-insecure, whereas 5.7% were vulnerable to food insecurity [63]. Earlier, in 2010, the prevalence of food security was higher (97.5%) among the studied households (n = 13,888), while food insecurity was found in 0.3%, and 2.1% were vulnerable to food insecurity [64]. These results may reflect a negative influence of the COVID-19 pandemic on food security status. Unfortunately, this effect could not be confirmed due to the lack of studies on food security status among the general population in Jordan shortly before the pandemic.
However, there are other studies of pre-pandemic food security status focused on specific groups such as women, vulnerable Jordanian communities, and children. A study of women in Northern Jordan (n = 500) found that 67.6% of the interviewed sample were from food-secure households, and 32.4% were from food-insecure households [65]. Further, a survey of 4241 vulnerable Jordanian households found a prevalence of food security (30%), risk of food insecurity (59%), and food insecurity (11%) [66]. A lower prevalence of food insecurity was found in 2008 by WFP [66]. Among 3000 vulnerable Jordanian households, 72.7% were food-secure, 19.2% were at risk of food insecurity, and 7.9% were food-insecure. Additionally, food security status was assessed among children (12-16 years) in Northern Jordan (n = 679), and it showed that the food security among children was 78%, while the mild and severe food insecurity were 15.3% and 6.8%, respectively [67]. Moreover, several studies have focused on the food security status of refugees [68][69][70][71]. The results of the current study could not be compared with the results of these studies due to the variety of the studied groups.
The study findings indicate the need for scaling up the implementation of national strategies to achieve food security by 2030 through addressing, holistically, all aspects of food security and adopting appropriate and resilient food systems, especially with the impact of the ongoing conflict in Ukraine, which is heightening concerns about the impact on food security worldwide. We do recommend that the decision-makers and healthpromotion professionals and institutions in Jordan plan interventional programs to reduce the consequences and magnitude of food insecurity, with an increased focus on social protection, care and nutrition services, and governance in Jordan, aiming for scaling up food security and food systems interventions, including the following recommendations: 1.
Ensure the availability of food at the national, household, and individual levels, with the following sub-objectives: A Achievement of the maximum potential of local food production B Provision of a sufficient and stable supply of imported items C Improvement of regional collaboration and integration of the different aspects of food security D Reduction in food loss and waste, and enhanced food safety 2.
Optimize the utilization and stability of food, with the following sub-objectives:

Study Strengths and Limitations
This study updates the data about food security status, its determinants, its effect on food consumption patterns, and the related household coping strategies. It is important to study these variables periodically, especially in low-and middle-income countries. The importance has increased considering the spread of COVID-19 and its economic consequences during the lockdown and beyond. In the present study, one of the main limitations was the young median age of the respondents, which may have occurred due to the use of social media as the distribution tool. Besides, online self-reporting questionnaires may lead to bias due to over-or under-estimation of the questions. However, the sample size, which was almost twice the size needed, is one of the main strengths of the present study, in addition to the fact that the participants in the current study were recruited from several regions in Jordan, and a validated questionnaire for measuring food security status was used.

Conclusions
Although the overall food security prevalence seems to have improved after the lockdown, around 16% of the participants are still food-insecure. Food insecurity was significantly associated with low educational level, being divorced or widowed, having large family size, a low number of the household members being employed, not having health insurance, and having a rented residence. Besides, most of the sample reported two or fewer meals consumed per day and a frequency of three days or less intake of almost all the main food groups per week. As expected, food insecurity was significantly higher among the participants who skipped meals and food groups, reported a "less than usual" number of meals, and had lower food consumption scores. Moreover, significant associations were found between food insecurity and food-based coping strategies such as eating cheaper foods, eating fewer meals or small amounts of food, and adults eating less food to spare it for their children. These results are consistent with results from numerous studies conducted worldwide, and they further underscore the impact of the COVID-19 pandemic on food consumption patterns, especially among food-insecure households, which may occur due to the loss of food accessibility, and hence decrease the quantity and quality of food consumption. These findings suggest severe impacts on people's nutritional, mental, and social health; therefore, there is a need for implementing national strategies to increase food access, such as stabilizing the prices of food products, providing employment opportunities, supporting small businesses, and increasing health and nutrition awareness. It is recommended to conduct further studies among other groups susceptible to food insecurity, such as people with chronic diseases, children, and refugees.